Police misidentifying drivers’ ethnicity when writing tickets, to avoid the appearance of racial profiling.
The Data Newsletter
BY BROOKE STEPHENSON

Welcome back, data nerds.
Today we’re digging into a case of missing data that reporters initially found inexplicable ... and suspicious.
Jefferson Parish is a community of 440,000 near New Orleans. Eighteen percent of the parish is Hispanic. But ProPublica reporters found that between 2015 and 2020, of the more than 73,000 traffic tickets the Jefferson Parish Sheriff’s Office issued between 2015 and September 2020, deputies identified only six of the cited people as Hispanic.
Six. In five years.
The notion that police pulled over only six Hispanic people out of a Hispanic population of more than 79,000 in five years is, well, pretty unbelievable.
Reporters were left thinking, “Well, that’s weird. But what can we really say about it?”
(Quick note about race and ethnicity before we go on: While Hispanic is an ethnicity, more than 80% of law enforcement agencies use it as a race when collecting information from drivers during traffic stops, according to a sample of 69 departments studied by one expert on racial profiling.)
Generally when we do data reporting, the work boils down to contextualizing numbers, and showing how a figure compares to what you might expect. For example, Black people make up 27% of Jefferson Parish’s population, so you might expect them to account for a similar proportion of all traffic citations. Instead, Black drivers account for 40% of all traffic citations. They are cited at a rate 1.5 times their population size. For reporters investigating bias, those numbers are a good place to start.
At first glance, the six citations of Hispanic drivers could imply that Hispanic people don’t get pulled over in Jefferson Parish very often. But the Jefferson Parish Sheriff’s Office has received multiple grants from the Department of Homeland Security to find and deport undocumented people, and the parish’s Hispanic residents told reporters stories of being pulled over frequently. So the most obvious explanation — that the data was accurate — looked unlikely.
These six citations of Hispanic drivers implied something different: missing data.
Reporter Richard Webster remembered a story by KXAN-TV about Texas troopers ticketing Hispanic drivers as white, which made it impossible to hold the police accountable for potential racial profiling. He thought something similar might be going on here.
But how could they prove it?
Data reporter Irena Hwang started by examining the dataset of white drivers.
She pulled out the most common last names. Among the ones that appeared the most often were Rodriguez, Martinez, Hernandez, and Garcia.
Those were the exact same last names that showed up most often in the dataset of white Texas drivers KXAN looked at.
Hwang then pulled census data and confirmed that people with the last names Rodriguez, Martinez, Hernandez and Garcia were more than 90% likely to identify as Hispanic. Another popular Hispanic name, Lopez, also appeared in the top ten last names of most frequently stopped Jefferson Parish drivers.
That seemed to indicate flaws in the records of identification of drivers in Jefferson Parish. But Irena kept working to zoom in further on the problem.
Irena took all the names that appeared in the dataset of white drivers, and looked them up in recent census data. If the name belonged to a Hispanic person a majority of the time, Irena made a note of that name.
She then started doing calculations.
Let’s say census data indicates that in the US, 90% of people with the name “Garcia” are Hispanic.
Irena ran a simple calculation:
(Number of people named “Garcia” in the driver database) x 90% = the likely number of people named Garcia in the “white” dataset who are actually Hispanic
By adding the estimated numbers of Hispanic people for each majority-Hispanic name in the database, Irena concluded that about 10% of drivers who are identified as white by Jefferson Parish officers are actually Hispanic. That’s about 7,400 drivers — and she thinks it’s an undercount because census data about ethnic and racial identity is available for only some of the names.
Although data can’t show why deputies are misidentifying motorists, accurate numbers are essential for understanding whether local law enforcement is illegally racially profiling drivers.
One of the experts the reporters spoke to, Frank Baumgartner, a political science professor at the University of North Carolina at Chapel Hill, put it like this: “If everybody’s white, there can’t be any racial bias.”
The Jefferson Parish Sheriff’s Office did not respond to emails and phone calls requesting comment.
Reporters did try to examine whether Hispanic drivers were being stopped disproportionately, but without accurate data collection, that was impossible. There were too many names in the dataset that, unlike Garcia or Rodriguez, could not be assigned to one race or ethnicity.
But local activists and residents help paint a picture of what policing is like for Hispanic people in Jefferson Parish.
Rachel Taber, a local immigration activist, said she has witnessed sheriff’s deputies cruising predominantly Hispanic neighborhoods, stopping people randomly, and then arresting them in coordination with U.S. Immigration and Customs Enforcement agents.
Another local activist, who asked to remain anonymous due to his immigration status, said, “I’m white-knuckled on the steering wheel, just making sure I’m always going the speed limit and that everything on my car is perfect. Because for any tiny thing, you just don’t know how the police might react and what could happen.”